Help! Brundle explained rev glitch and missed commentary

[Mr Satay]

Hello, during the Turkish GP Brundle mentioned what he thought was responsible for the "glitch" you hear on on-board shots of cars when reving. He said something about a "cell" but thanks to Channel 10s (australia) AWESOME commercial breaks, they cut right into one mid sentence.

Can anyone remember what he said was the reason for the glitch? I really want to know as its damn annoying!

[Giblet]

I would imagine the came to the same conclusion I did, that the cars need to broadcast their compressed digital feeds to the various transponders around the track.

Much like a cell phone system, the transponders/receivers have a limited distance of transmission, so there are more than one, and each on has it's own "cell".

As the car moves from cell to cell, the signal needs to be switched over from one receiver to the next, and during this switchover, which by nature will not be instantaneous, there is a period during the final output to the network where the sound speeds up and slows down via time compression as the buffered signal lines it's timecode back up with real time.

This phenomenon is easily seen in digital videos watched on a computer. Sometimes the video skips or pauses and goes really fast to catch back up to the audio.

[modbaraban]

It's funny, though, that we didn't notice much of this before. Not that there's KERS, many started wondering, some have automatically connected that to KERS usage.

[paused]

Only just a guess, but perhaps they have changed the system recently and combined with the introduction of Kers gives the impression it was Kers related. After all there were quite a lot of problems with transponders last year. Here is a link to a blog about the timing system failure during the 2008 German GP

http://vee8.doctorvee.co.uk/2008/07/22/ ... structure/

[Giblet]

I always get the speedTV coverage, which is just the ITV feed and them talking over it. Same as TSN here.

I used to download SkyTV (I think) in car footage from the races, with no commentary in widescreen, and have it running along the regular race footage on a second monitor, to truly nerd out.

Is that footage still available, and if so, it would be a great way to analyze repeated instances at locations.

Time to troll the intertorrents.

[mike]

i think its the gear selectors in a seamless shift gear box.
they refer to it as cells, it will be more apparent if the revs are less since the time spent over driving is increased, and it have nothing to do with KERS.

[Giblet]

It's nothing to do with the car, it is a digital audio glitch.

I downloaded the in car footage, and I have been looking, and like i also noticed at the last race while looking frame by frame , every time the glitch happens, it is accompanied by a small video glitch, and it happens at the same place at each lap. The repetition makes it not a coincidence.

It doesn't matter the car, or the fact the drivers hands are often doing nothing at all like in the last race, where the glitch, both video and audio, were about a car length past the star/finish line.

[paused]

It's nothing to do with the car, it is a digital audio glitch.

I downloaded the in car footage, and I have been looking, and like i also noticed at the last race while looking frame by frame , every time the glitch happens, it is accompanied by a small video glitch, and it happens at the same place at each lap. The repetition makes it not a coincidence.

It doesn't matter the car, or the fact the drivers hands are often doing nothing at all like in the last race, where the glitch, both video and audio, were about a car length past the star/finish line.

Giblet, actually your post has just made me realised that the "cell" explaination is not correct. If that were the case you would notice this in other parts of the race track which do not seem to be the case. As you rightly point out this seems to occur just after the straight finish line only.

Therefore is much more likely that the glitch is the result of some sort of ECU download that occurs once per lap after crossing the finish line. After all the transponder and FOM systems are all connected to the ECU. Does anyone know what data the FIA download on a lap by lap basis?

[modbaraban]

As you rightly point out this seems to occur just after the straight finish line only.

I remember hearing one on the back straight in Turkey.

[enkidu]

Hello, during the Turkish GP Brundle mentioned what he thought was responsible for the "glitch" you hear on on-board shots of cars when reving. He said something about a "cell" but thanks to Channel 10s (australia) AWESOME commercial breaks, they cut right into one mid sentence.

Can anyone remember what he said was the reason for the glitch? I really want to know as its damn annoying!

I must admit when he said about this I wondered if he had been reading f1technical lol.

He just said what a guy above said, the receivers are setup in cells like mobile phone are setup. Hence why in america they call them cell phones. As the cars pass from one cell to another its causing a sound glitch.

[tarzoon]

The problem becomes more evident in the straights because of different effects aggravated by speed, such as Doppler.

When the car approaches a new cell, the sound and image goes faster. The opposite happens when it moves away. In the image it's barely noticeable, but in sound it's there for everyone to listen.

In other words, think of an ambulance siren at 300+ kph

[tommylommykins]

The problem becomes more evident in the straights because of different effects aggravated by speed, such as Doppler.

When the car approaches a new cell, the sound and image goes faster. The opposite happens when it moves away. In the image it's barely noticeable, but in sound it's there for everyone to listen.

In other words, think of an ambulance siren at 300+ kph

This is certainly wrong. The speed at which the signals travel hardly changes at all as a proportion of their full speed. When they go at hundreds of thousands of meters per second anyway, an extra hundred or so from the cars would make no difference.

Anyway, because these signals are digital, any minor doppler effect would not change the pitch of the sound.

[Saribro]

Doppler

Radio signal : 300 000 000 m/s = 1 080 000 000 km/h
Car : 300km/h
Difference : 2.7e-7 (or 0.000027%) aka, utter insignificance.

The ECU "download" seems plausible. We've had a similar issue with a recent project where we had a microcontroller that had to multiplex it's single SPI port between an ethernet controller and another microcontroller. It was tricky to get the timing and duration of inter-microcontroller traffic right in order to not cause indeterminism and connection errors on the network end.

[tarzoon]

Radio signal : 300 000 000 m/s = 1 080 000 000 km/h
Car : 300km/h
Difference : 2.7e-7 (or 0.000027%) aka, utter insignificance.

The ECU "download" seems plausible. We've had a similar issue with a recent project where we had a microcontroller that had to multiplex it's single SPI port between an ethernet controller and another microcontroller. It was tricky to get the timing and duration of inter-microcontroller traffic right in order to not cause indeterminism and connection errors on the network end.

Yep, you're right. Didn't even bother calculating

Still, I think there might have to do with speed, data compression and error correction. Anthony Davidson explained it on the 37th minute.

http://news.bbc.co.uk/sport1/hi/motorsport/7981604.stm